Reply

Abstract

Wang et al. summarized published reports and reached a result to support a potential negative association between thiazolidinediones (TZD) therapy and risk of liver cancer among type 2 diabetes patients. Evidence suggests that nonalcoholic fatty liver disease (NAFLD) is an independent risk factor for hepatocellular carcinoma (HCC). Several clinical trials demonstrate that TZD therapy may lead to improvements in certain metabolic and histological parameters in patients with NAFLD.1 However, the long-term RECORD and PROactive trials failed to find any difference in liver cancer occurrence between either rosiglitazone or pioglitazone and the active comparison group, probably because patients with chronic liver disease were excluded from the trials. Southeast Asia, including China, Taiwan, and Hong Kong, is the endemic area for chronic viral hepatitis. Our case-control study2 and Lai et al.'s cohort study,3 both using Taiwan's National Health Insurance Research Database, demonstrated that the use of TZD was associated with a reduced risk for HCC. There has been a concern that several methodological limitations, in particuluar, the inability to control for important clinical and behavioral factors, might affect the validity of observational studies that were based on large healthcare claims databases. In a recently published report, Yang et al. analyzed the Hong Kong Diabetic Registry and found that TZD was associated with lower overall cancer risk in a dose-response manner after controlling for potential confounding factors.4 However, the small number of TZD users did not allow a further exploration of the association between TZD use and site-specific cancers. Although some evidence suggests a potentially beneficial role of TZD against liver cancer, many questions need to be answered. For patients with NAFLD, does TZD therapy that leads to decrease in aminotransferace, reduction of steatosis, hepatocyte ballooning, or fibrosis result in decreased HCC risk? Among high-risk patients with chronic viral hepatitis B or C, does TZD therapy improve liver pathology and further reduce liver cancer risk? What is the potential role of TZD along with antiviral therapy? Is there any possible effect modification by viral genotypes?5 Is there any difference in risk reduction between rosiglitazone and pioglitazone? Are there any clinical, metabolic, or pathological factors that can predict response? We agreed with the researchers that all these questions have to be answered by well-designed studies with meticulously collected information on metabolic, biochemical, virological, and histological parameters. Randomized, controlled trials with intermediate outcomes may provide more solid evidence in support of the potential beneficial effects associated with TZD. Chia-HsuinChang M.D., D.Sc. Jou-Wei Lin M.D., M.P.H., Ph.D. Mei-Shu Lai M.D., Ph.D. Lee-Ming Chuang M.D., Ph.D. Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan Cardiovascular Center, National Taiwan University Hospital Yun-Lin Branch, Dou-Liou City, Taiwan Institute of Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan